How NPA Strategy AI Decides Between Settlement, SARFAESI and Write-Off

The decision between settling an NPA, invoking SARFAESI, or writing it off is the highest-value judgment call in the recovery function — and the one most often made on instinct, precedent, and relationship pressure rather than evidence. The NPA Strategy AI makes it on net present value, collateral realisation probability, borrower capacity, and legal timeline modelling — every time, for every account, with a documented rationale that survives regulatory scrutiny.

Why the Decision Is So Consistently Made Poorly

In most lending institutions, NPA resolution strategy emerges from a combination of recovery officer intuition, the loudest voice in the committee room, and the institutional bias of whatever strategy produced good results in the last cycle. A recovery head who built a career on SARFAESI proceedings tends to invoke SARFAESI more frequently than the economics justify. An institution that wrote off aggressively during a previous stress cycle tends to write off more readily than it should. A settlement culture produces settlements that undervalue the institution's legal position.

None of these biases are malicious — they are the product of experience applied too broadly. But they produce systematic value destruction: accounts settled too cheaply when SARFAESI would have recovered significantly more, accounts sent into multi-year legal proceedings when a swift settlement would have recovered more on a net-present-value basis, accounts written off with residual recovery potential that was never pursued.

The NPA Strategy AI removes the bias by making the decision on a consistent, evidenced analytical framework — one that is applied identically regardless of the account's relationship history, the recovery officer's preferences, or the committee dynamics of the day.

The Decision Framework: Six Input Dimensions

The NPA Strategy AI evaluates every NPA account across six dimensions before generating a resolution recommendation. No single dimension is determinative — the recommendation emerges from the intersection of all six, with different weights applied based on the account's collateral type, borrower category, and outstanding exposure.

Four Account Profiles: What the AI Recommends and Why

The NPV Comparison the AI Produces for Every Account

For every NPA account where the recommended resolution path is not obvious, the NPA Strategy AI produces a full NPV comparison across all three primary resolution strategies — modelling the expected recovery quantum, the time to recovery, the all-in cost of each path, and the probability-weighted outcome. The comparison is account-specific: it uses the actual collateral valuation, the applicable legal timeline for the jurisdiction and court, and the institution's own historical recovery rate data for similar accounts.